The Root Axiom · Core Proof and Derivation
The axiom. RA: ∀x ∈ 𝕌, ∃x ⟹ ΔE_k(M_x) > 0. To exist is to actuate: any existent x forces a strictly positive kinetic expenditure in any substrate M_x that registers it.
I. Mathematical Proof of RA
1. Definitions. x is operationally existent iff some registration of x exists: a physical process in some substrate M_x whose final state is distinguishable from its initial state conditional on x, so that the universe with x and the universe without x differ at that register. Distinguishability of states is orthogonality in the substrate’s state space. ΔE_k(M_x) denotes the energetic resource of the registering transition: the energy spread ΔE and the mean energy above ground ⟨E⟩ − E₀ are its two measures.
2. Lemma (Frozen Substrate). If ΔE = 0, the substrate state is an eigenstate of its Hamiltonian, so ψ_t = e^(−iEt/ℏ) ψ₀ and |⟨ψ₀|ψ_t⟩| = 1 for all t. A zero-spread substrate never reaches any state distinguishable from its initial one. Nothing is ever registered by it. [⟀]
3. Theorem (Kinetic Floor). Any transition to an orthogonal state takes time
τ ≥ πℏ / (2ΔE) (Mandelstam-Tamm) and τ ≥ πℏ / (2(⟨E⟩ − E₀)) (Margolus-Levitin).
Proof of the first bound. For any observable A, Robertson gives |d⟨A⟩/dt| = |⟨[H, A]⟩|/ℏ ≤ 2 ΔH ΔA / ℏ. Take A = |ψ₀⟩⟨ψ₀|, so ⟨A⟩ = P(t) = |⟨ψ₀|ψ_t⟩|² and, A being a projector, ΔA² = P(1 − P). Hence |dP/dt| ≤ (2ΔE/ℏ)·√(P(1−P)). Substituting P = cos²θ: |dθ/dt| ≤ ΔE/ℏ. Orthogonality is the arc θ: 0 → π/2, so τ ≥ (π/2)·ℏ/ΔE. [⟀]
The two bounds bind every actuation, reversible or irreversible: the floor is an energy-time resource and requires no dissipation.
4. Theorem (Dissipative Floor, the irreversible subclass). Erasing one bit merges two distinguishable record states into one: ΔS_record = −k_B ln 2. Total entropy is nondecreasing, so the environment absorbs ΔS ≥ k_B ln 2, hence heat Q ≥ k_B T ln 2 per erased bit (Landauer). The floor is measured (Bérut 2012). [⟀]
5. Theorem (RA). ∀x ∈ 𝕌: ∃x ⟹ ΔE_k(M_x) > 0.
Proof. Suppose ΔE_k(M_x) = 0 across every register: every candidate registering substrate sits at zero spread and zero energy above ground. By the Lemma each such substrate holds unit overlap with its initial state for all time; by the Kinetic Floor no orthogonal excursion completes in any finite time. No registration of x exists, so x fails operational existence by Definition 1. Contrapositively, every existent forces at least one registering transition, and every transition carries the strictly positive floor of Theorems 3 and 4. [⟀]
II. The RA Derivation
1. The parse. RA carries exactly three irreducible components: A₁ the existent (∃x), A₂ the actuation (ΔE_k > 0), A₃ the binding (⟹). Delete A₁: an unbound predicate, a cost with no bearer. Delete A₂: a conditional with no consequent, existence asserted at no price. Delete A₃: two adjacent assertions and no law, the mandate gone. Each deletion destroys the claim and no slot is recoverable from the other two. The three lexicons, quantifier, energetic, connective, are pairwise disjoint. Three slots, no more, no fewer.
2. The axes and the composition law. The slots map to three verification axes: A₁ → V_F, A₂ → V_E, A₃ → V_ER. Iterated audit imposes three closure requirements on the algebra the axes generate: bracketing invariance of repeated audits (associativity); nonzero warrants never compounding to zero (no zero divisors); a linear evidence register with unit ground (real algebra with identity).
3. Theorem (Unique Carrier). The algebra closing three orthogonal axes under these requirements is ℍ, and exactly ℍ.
Proof. Scalar Exit: in a normed algebra the rank equation gives u² = 2⟨u,1⟩u − ‖u‖² for every u; a pure unit satisfies u² = −1. Self-composition of any axis exits the axes entirely and lands on the scalar line: a closed system of axes must carry a ground slot. Fertile Orthogonality: for orthogonal pure units, ‖uv‖ = 1 by norm multiplicativity, Re(uv) = −u·v = 0 makes uv pure, and uv is orthogonal to u and to v; the minimal multiplicatively closed span of two orthogonal axes is {1, u, v, uv}, dimension four. Frobenius: the associative real division algebras are ℝ, ℂ, ℍ, with axis counts 0, 1, 3. The parse demands three; only ℍ supplies three; the octonions forfeit associativity and the sedenions forfeit integrality through zero divisors. The triad is Im ℍ, the ground is ℝ, and the demand matches the supply exactly. [⟀]
4. Theorem (The RA Line). Z(ℍ) = ℝ, and ℝ is the unique line of ℍ fixed pointwise by every automorphism.
Proof. Write q = a + bi + cj + dk. Commutation qi = iq forces c = d = 0; commutation qj = jq forces b = 0; so the center is ℝ. Every automorphism of ℍ is inner, q ↦ rqr̄ with |r| = 1, fixing ℝ pointwise and acting on Im ℍ as SO(3); the standard SO(3) representation is irreducible over ℝ, so no line of Im ℍ is invariant. [⟀]
Conjugation, the unique anti-involution with (pq)̄ = q̄ p̄, splits ℍ = ℝ ⊕ Im ℍ into its +1 and −1 eigenspaces. The axes invert under reversal of audit order; the ground does not. The axiom occupies the one register of the verification algebra that no frame change and no reordering can touch.
5. Theorem (The Return). For the composed triad w = q̂_F q̂_E q̂_ER of pure units, the projection onto the RA line is
λ = Re(w) = (w + w̄)/2 = ½ (q̂_F q̂_E q̂_ER − q̂_ER q̂_E q̂_F),
since w̄ = (−q̂_ER)(−q̂_E)(−q̂_F) = −q̂_ER q̂_E q̂_F. Hence: contact with the RA line, λ ≠ 0, holds iff the forward and reversed audits differ; collapse, λ = 0, holds iff the audit is palindromic, q̂_F q̂_E q̂_ER = q̂_ER q̂_E q̂_F, in which case w is a pure unit and w² = −1; the full Return, w = ±1, is reversal exactly negating the composition. At the right-handed landing: ijk = −1, kji = +1, λ = ½(−1 − 1) = −1. GOL ⟺ λ ≠ 0. [⟀]
6. The instrument obeys its axiom. In the pipeline, each axis row carries d_a = ‖m̃_a‖²/(N−1), its kinetic mass. A zero-variance row, evidence in which nothing ever actuated, gives d_a = 0 and det(G) = d_F d_E d_ER · det(R) = 0 before geometry is consulted: an axis that does not actuate cannot enter verification. And the audit itself is a transition in the auditing substrate, so every verdict pays the Part I floor. Existence of a verdict is itself an instance of RA.
Conclusion. Existence costs actuation, with the cost floored by ℏ at every transition and by k_B T ln 2 at every erasure. The cost statement parses into exactly three slots; the slots force exactly three axes; the axes close into exactly one algebra; that algebra owns exactly one frame-invariant line, the scalar center, the axiom’s own. The verdict functional is the projection back onto that line: every verification of anything terminates where existence begins. RA is the line. M_seal is the touch. [⟀]
The Geometric Orthogonal Lock · Core Derivation
Setup. X = S² × S² × S². T is the 3 × 3 matrix with unit rows v_F, v_E, v_ER. R = TTᵀ is the unit-diagonal Gram matrix. Each row reads as a pure quaternion q̂. The verdict functional:
λ = Re(q̂_F q̂_E q̂_ER) = −(v_F × v_E) · v_ER = −det(T), det(R) = det(T)² = λ².
I. Mathematical Proof of the Seal
1. Anatomy of the composition. Let w = q̂_F q̂_E q̂_ER. Norm multiplicativity gives |w| = 1 identically. Expanding (−v_F·v_E + v_F×v_E)·q̂_ER with the triple-cross identity:
w = λ + Im(w), Im(w) = (v_F·v_ER)v_E − (v_E·v_ER)v_F − (v_F·v_E)v_ER.
2. Bounds and equality cases. R is a Gram matrix, so det(R) ≥ 0, with equality exactly at linear dependence. |λ| = |Re(w)| ≤ |w| = 1, so det(R) ≤ 1. If Im(w) = 0 then |λ| = 1, the rows are independent, and independence forces every coefficient in the closed form of Im(w) to vanish: the triad is orthonormal, R = I. Conversely orthonormality kills every inner product in Im(w). Hence
det(R) = 1 ⟺ orthonormal ⟺ w = ±1, det(R) = 0 ⟺ λ = 0 ⟺ w pure unit ⟺ w² = −1.
The two saturations of |w| = 1 are the scalar and the pure axis. Nothing else is available at the boundary.
3. The landing and its gauge. At orthonormal T ∈ O(3), det(T) = ±1. Right-handed: det(T) = +1, λ = −1, the Hamilton relation ijk = −1. Left-handed: λ = +1. Under the diagonal action v ↦ Av, A ∈ O(3): T ↦ TAᵀ, λ ↦ det(A)·λ; under axis relabel σ: λ ↦ sgn(σ)·λ; quaternionically the rotation case is conjugation, Re(rwr̄) = Re(w). The invariant of the seal under the full O(3) × S₃ action is
det(R) = λ² = 1,
and the seal verifies at the determinant register: det(R) = 1 > ε.
4. The RA line. λ is the orthogonal projection of w onto ℝ = Z(ℍ), the line named for the Root Axiom. Contact: λ ≠ 0. Full Return: w = ±1, the composition collapsed onto the line itself. Exact collapse: w² = −1, the composition a pure axis with zero projection. [⟀]
II. Derivative and Stability of the Seal
1. The rotational derivative, computed once for all configurations. Jacobi’s formula: d/dt det(T) = tr(adj(T)·Ṫ). A common rotation of the rows, v̇ = ω × v with Ω = ω× skew, is Ṫ = TΩᵀ in row convention. Since adj(T)·T = det(T)·I for every square matrix,
dλ/dt = −tr(adj(T)·TΩᵀ) = −det(T)·tr(Ωᵀ) = 0 at every T ∈ X, locked or not.
At the seal specifically, adj(T) = det(T)·Tᵀ = −λ·Tᵀ, and the same zero falls out. This is frame invariance, the conjugation law in differential form. It holds on the entire configuration space.
2. Criticality at the seal. The Riemannian gradient component at v_F is grad_F λ = −(I − v_F v_Fᵀ)(v_E × v_ER), cyclically for the others. At an orthonormal triad, v_E × v_ER = ±v_F, which the projection annihilates. Therefore dλ = 0 at the seal in every tangent direction, rotational or not: first-order λ is identically blind there, for breaches and non-breaches alike.
3. The critical dichotomy. For f = λ² = det(R), grad f = 2λ·grad λ. Either λ = 0, the collapse variety, where the gradient vanishes identically and f = 0, or λ ≠ 0 and criticality forces v_E × v_ER ∥ v_F and cyclically, hence pairwise orthogonality and f = 1. The critical values of the verdict functional on X are exactly {0, 1}. The value-1 stratum is O(3), two components by orientation. No critical configuration exists at any intermediate value.
4. Second order, where stability lives. At T = I, pair the six tangent tilts by coordinate plane. In the e₁e₂ pair, with v_F = (e₁ + a e₂)/√(1+a²), v_E = (e₂ + b e₁)/√(1+b²), v_ER = e₃:
f(a, b) = (1 − ab)²/((1+a²)(1+b²)) = 1 − a² − 2ab − b² + O(degree 4),
block Hessian [[−2, −2], [−2, −2]], eigenvalues −4 on (1, 1) and 0 on (1, −1), the null direction being the rotation about e₃. Cross-plane second derivatives vanish, so the full spectrum is
{−4, −4, −4, 0, 0, 0},
kernel exactly the common-rotation directions, the tangent space of the stratum; equivariance transports this to every lock point. Hence for a tangent perturbation ξ with transverse part ξ⊥:
δλ = 0 at first order, det(R) = 1 − 2‖ξ⊥‖² + O(‖ξ‖³).
The breach signal is second order in det(R), at uniform rate −4 transverse to the stratum. First-order λ carries none of it.
5. The basin. Run gradient ascent Ṫ = grad f on X and set U = 1 − det(R). Then U̇ = −‖grad f‖² ≤ 0, with U = 0 exactly on the lock stratum and U = 1 exactly on collapse. X is compact, f ascends to a limit f̄, the ω-limit set is invariant and critical at value f̄ ∈ {0, 1}. If f(0) > 0 then f̄ = 1 and the trajectory converges into the lock stratum; convergence is to a single point because f is real-analytic (Łojasiewicz) and the stratum is Morse-Bott nondegenerate by step 4. On {λ = 0} the gradient is identically zero: collapse is stationary, of measure zero, and monotonicity keeps every trajectory with f(0) = c > 0 inside {f ≥ c}, away from it forever. The basin of the lock is X ∖ {λ = 0}: open, dense, full measure.
Conclusion. Under the gradient flow of its own verdict functional, the lock stratum is the unique attracting critical manifold, uniformly strict at −4 in every direction that leaves it; the collapse variety is the unique repelling stratum, stationary and of measure zero; the only critical values are 0 and 1, so an ascending trajectory has exactly one kind of place to end. The rotational zero is symmetry. The all-directions zero at the seal is criticality. The −4 is stability. [⟀]